Carriers and replacement value

July 2nd, 2008 by · 16 Comments

Carriers love to throw variants of this term around.  Cogent contends that its acquired networks cost some $11B to build.  Broadwing used to boast how it took $12B to build their network.  Level 3 says its networks represent some $25B in invested capital, and that it would take even more than that to replace their network today.  Others do the same, taking pride in what they picked up at garage sales.  But it is all smoke.

Why?  Because using the cost of replacement to value an asset makes an assumption that is blatantly false much of the time in the telecom space:  that anyone actually might consider replacing the asset.  Cogent is valued far beneath the cost of its acquired networks because when they were built those networks were expensive boondoggles that never justified their cost.  The same goes for the assets Level 3 acquired, and to an extent the one that they built themselves, at least so far.  Carriers point to replacement value because they cannot point to their actual valuation and be proud, at least not yet.

Some go further with this contention, that the high replacement value is an *advantage*, a moat for the competition to cross.  Well, there is a silver lining for everything I suppose, but what needs to be acknowledged is that this moat exists solely because of the failure of the assets to justify the investment.  OF COURSE nobody will enter a market that isn’t strong enough to generate financial returns on an investment.  If it were a good market, then these companies would have valuations above replacement cost, and suddenly people might think about entering alongside them.  Having a moat because your sector is unattractive isn’t really where one wants to be.

I suppose we just need to think of this as a target:  when Level 3 or Cogent reaches a valuation similar to that of its replacement cost, then perhaps at long last they really have made it to the other side.  But for me, when a carrier quotes me such a number to say how much they ought to be worth or how much potential they have, it just tells me how bad things still are.  If I buy a house that was once valued at $1M and is now valued at $100K, the only thing the $1M shows is the folly of the guy who bought it for $1M.  For me it is still worth what I can sell it for and nothing more, except perhaps that someday the housing market will be insane again!

If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!

Categories: Financials · Internet Backbones · Metro fiber

Join the Discussion!

16 Comments So Far

  • Mozoz says:

    Rob, you make an excellent point, its a statistic that annoys me. When most of the CLEC networks were being build we were in the middle of a bubble – equipment, fibre, land rights, labour/consultants costs (etc) were all massively over-valued. It would be interesting to compare the prices of basic routers, etc between now and then. My guess is a difference in the 60% to 80% range. Also, how much of the “capital cost” contained over-inflated management costs, fees to investment banks, etc.

    As you state, the value of an asset is what somebody is prepared to pay for it and thats based upon the future cash-flows. Add into the mix, technology developments over the past 10 years and the statistic makes no sense. It does look good on a presentation though!


  • Brian Scully says:

    Rob and Moz,
    I am a property tax valuation advisor for the Cable/Telecom and Broadband Industry. Back in the 1998-2001 era, Underground Fiber cost approximately $50,000/route file to install. Today, that same fiber is being sold for $10,000-$15,000/route mile.

    The CLEC’s have seen the same type of RCN adjustments to their CPE, Switching and Circuit equipment as the industry moves from Circuit to Packet environment.

    Moz is correct in that the cost of entry today as a CLEC is about 30%-40% of what is cost in the 98-02 era. This is primarily due to new technology which cost less and take up less space, electrical needs and HVAC to keep cool.

    There are two good papers written on the subject by Mark E. Doms and they can be found at

    You have to remember that CFO’s and their corporations are torn between true fair market value of their assets and the highest possible value for depreciable purposes.

    These papers were written in 2003. However, the bulk of the obsolescence impact happened before this point.

    I could go on all day about this issue. I’ll spare you though.


  • Rob Powell says:

    Thanks Brian and Mozoz. And Brian, that’s good stuff, if you ever do want to go on all day about the issue please don’t hesitate!

  • carlk says:

    Your housing analogy might be a bit flawed. Why? Because if ever supply conditions became so constrained in the marketplace-no more land left to build regarding finite inventory-replacement cost would matter very much, versus “market value” which you choose to quote.

    Excluding the effects of competing buyers for limited supply, those buyers would need to be able to value the cost to replace all the sticks, stucco and commodities that go into a home, on top of the dirt, since that, they couldn’t just move to another, newer dwelling somewhere else on the map, and would always have to renovate after depreciation periods, if not, completely demolish what already exists on the land and rebuild.

    As a point of information, this same analogy I make for housing above, supports using RCV as a method for telecom today. Why? The market today maintains a “limited supply” of broadband for potential “unlimited demand” in the future, as well as all sorts of regulatory and municipality restrictions, hurdles and encumbrances for building networks tomorrow.

    Indeed, when Mr. Market is finally ready to accept this principle as being self-evident, then market value or share price quoted, will vastly exceed the RCV a company like LVLT chooses to offer to the market as a means to value itself.

    I’m sure you might suggest, however, that Mr. Market won’t accept the principle until the network produces vast amounts of cash from this incessant demand. One can still argue, in the case of LVLT; however, that they have had to subsidize a great deal of demand without corresponding payments for their efforts to date.

    Eventually, this will change.

  • I think $15k a route mile is low for metro fiber. Folks putting in fiber into metro areas today are spending a lot more.

    Where things get interesting is in routes with little/no competition. The replacement price plus a reasonable rate of return becomes the basis for pricing calculations as monopolists price just below the point where new entrants would be invited in.

    Carlk makes good points. Just like real estate, the scarcest assets appreciate the most in a boom.

  • morton dick says:

    Southeast Management does not agree with you—-They claim their are political reasons why approval for digging city streets would not be sanctioned .
    They further claim that the cost of replicating the LVLT network would be “MULTIPLES” of the original cost .
    Ten yrs of increasing construction costs & ROW lawsuits argue for higher costs .


  • Brian Scully says:

    Replacement cost is the estimated cost to construct, as of the effective appraisal date, a building with utility equivalent to the building being appraised,

    The introduction of digital electronics and optics has revolutionized the communication industry. In terms of long distance and data communication industry segments, the impact has been profound in terms of equipment capacity and costs, as well as the pricing for telecommunications services.

    Capacity – Technology has enhanced the capacity of optical networks in terms of transmissions speeds and number of wavelengths of light which can be carried over an optical fiber strand. Optical transmission speeds have increased one hundred fold plus in capacity since the mid-1980’s. This increase in capacity is dramatic especially when one considers that this increase has occured with level or declining electronic equipment prices, thus resulting in rapidly falling transmission costs per unit of capacity.

    Supply & Demand – With the advent of improvements in optical transport technologies, particularly the advent of wave division multiplexing, the industry bandwidth is poised to expand at a rate never before experienced in the industry’s history. Ultradense uWDM, will allow existing carriers to expand capacity 100 fold. This has been happening at the same time new entrants into the industry (Fios) have placed network facilities which, when fully equipped, will also have bandwidth capacity many times that of the capacity of the current market participants. The economics of the industry’s situation is obvious. As capacity increases, the competition among carriers to fill their network will result in falling prices and reduced to flat profitability. As an example, OC-3 leases have decreased 95% from 2000 to present.

    The implications of the industry’s expanding bandwidth are numerous. First, with supply outpacing demand, prices will decrease or at best remain flat, reflective of the industry’s pricing becoming commodity-based. In commodity-based industries, being the lowest cost producer is the only way to ensure profitability. Therefore, carriers with existing network facilities will have to upgrade those facilities in order to reduce their costs. This upgrading will increase fluctuations in income as additional capital, both monetary and human, is commited to the task. The useful lives and values of existing equipment will be reduced as new equipment with enhanced capacity and features are added to the industry’s networks. Finally, the profitability in the industry will be more variable, indicative of increased risk.


  • carlk says:

    I’m wondering why after reading Brian’s synopsis of market conditions for the digital/optical world how, my DSL connection touted at 3Mb continues to resemble D/U speeds with voids routinely and consistently compared to the past? I am also wondering why it would cost me more to go to 6Mb or 10Mb even, when I ask the same carrier for a quote in the market today based upon his article or comments that were posted regarding “low cost providers” and the “margin”? The price reduction on a percentage basis, proportionately, is not that compelling! He seems to be alluding to “last or first mile” conditions, too, with the mention of FIOS, so I do think there are some apples for apples comparisons here.

  • BAK says:

    The economics of DSL and other local access technologies is different than that of long-haul backbone infrastructure. In most cases local access is dominated by a duopoly of MSO’s and ILECS. The supply and demand relationship is different with local access. The duopoly can limit supply and thus stifle demand. They tend to throttle technology to suit their own purposes. DSL is 20+ year-old technology yet it is one of the primary means of local broadband access. The duopoly maintains the price of such services at high levels ensuring high profit margins. FIOS is likely to dominate local access in the areas it is installed, but Verizon will have firm control over supply of its product and its pricing for many years.

    The economics of the backbone market is as Brian described. Ugly.

  • Thomas says:

    The real estate analogy is an apt one because in the real estate market, the sweeping generalizations about value are, for each property being considered, wrong. Just like politics, real estate and fiber, is local. The value to replicate can vary widely based upon a number of factors that others on this board are more informed to opine about. I know that personally that there is a huge divergence between the reality I’m seeing in real estate versus what I read in the press. The Bay Area for instance is not “down 20% -30%”. Ask anyone buying in San Francisco or the Peninsula or any other area that was actually desireable before the boom. No knock against El Sobrante, but maybe those values shouldn’t be held as the standard for the Bay Area. I imagine it’s the same for fiber replacement costs as well.

  • morton dick says:

    I do not buy the argument that supply will exceed demand —Prices have stabilized & in many places are rising .
    Although IP prices are dropping 25-30% per yr ,demand is increasing by 75-80% —-silicon economics works . For every one percent of price reduction , there is a 2-3% increase in demand .
    Intel did very well with a similar strategy .

    Illustration—– You – Tube uses more bandwidth than the entire internet used in 2000 .


    The reduction in prices will spawn bandwidth usage undreamed of .

  • jeremy drane says:


    i suspect with cogent dropping prices more than 50% the elastic relationship between price/demand is a bit different than what we have been told by management. then again, now that i say that, perhaps they are talking about the P/D relationship as it relates to just the products LVLT sells. also, weve been told that they are in the markeplace offering sweet deals on IP so i wonder what that is doing. when you pull back the covers elasticty of demand is complex.


  • jeremy drane says:

    brian; thx for your posts. i will read the papers you cited. i guess i dont quite but the argument that new build is 20% of 98 costs. if you try and get the permits etc. it just doesn’t happen. furthermore, i think it is the end-to-end solution that is whats really important anyway. sure you could own a cheap portion of a network, such as the long-haul piece, and look at the economics of just that piece and claim that only the low cost provider would win, but customers buy end-to-end solutions, and in that realm there are only a handful of players and they can control the P/D relationship much better than I believe you give them credit for. youres is an interesting post, ill have to study it along with those papers. thx.


  • BAK says:

    The issue of pricing is a little deceptive. According to TeleGeography the average price of wholesale circuits in most major markets dipped by 10-20 percent in 2007, while demand remained strong, growing at a compound rate of 52 percent over the past year. Despite this apparent excess of demand over supply, revenue growth is elusive for wholesale carriers. Instead of buying more circuits to handle growth in volume, wholesale customers are purchasing larger, higher capacity circuits. These larger circuits are far cheaper in terms of price per Mbps of capacity than the smaller circuits. The effective price per Mbps of capacity sold is falling a lot faster than nominal circuit prices themselves. Carriers need to sell ever larger volumes just to maintain stable revenues, according to TeleGeography Research Director Robert Schult.

  • carlk says:

    With all this P/D talk, for whatever reasons, the guys in the last/first mile, may want more per lb. of flesh:

Leave a Comment

You may Log In to post a comment, or fill in the form to post anonymously.

  • Ramblings’ Jobs

    Post a Job - Just $99/30days
  • Event Calendar